Data Table (skbio.table)#

This module provides support for interaction with data tables.

BIOM table#

Biological Observation Matrix (BIOM) is an efficient and versatile sparse table format designed for biological “omic” data types. It is the native table format in scikit-bio.

Table(data, observation_ids, sample_ids[, ...])

The (canonically pronounced 'teh') Table.

example_table

The (canonically pronounced 'teh') Table.

Table-like formats#

scikit-bio functions directly operate on various “table-like” formats, such as BIOM table, Pandas and Polars dataframes, NumPy array and AnnData objects, without the need for explicit format conversion. Read below on the specifics, nomenclature and usage of supported table-like formats.

Data augmentation#

Techniques for creating synthetic samples based on the current data and biological properties. Helpful for improving the accuracy and robustness of machine learning models.

phylomix(table, n, tree[, taxa, labels, ...])

Data augmentation by PhyloMix.

compositional_cutmix(table, n[, labels, ...])

Data augmentation by compositional cutmix.

aitchison_mixup(table, n[, labels, ...])

Data augmentation by Aitchison mixup.

mixup(table, n[, labels, intra_class, ...])

Data augmentation by vanilla mixup.